Re: [vpp-dev] vpp crash in counter

[hemant via lists.fd.io]

I fixed my plugin counter code and the VPP generic counter code's crash
below is fixed.  

 

Hemant

 

From: vpp-dev@lists.fd.io  On Behalf Of hemant via
lists.fd.io
Sent: Wednesday, March 24, 2021 9:49 AM
To: vpp-dev@lists.fd.io
Subject: [vpp-dev] vpp crash in counter

 

I did download latest vpp from gerrit today to check if the crash is
reproduced and it is. I have used a debugged image to debug the issue.  This
crash happens during testing of my plugin which is running IPv6 and the
plugin uses simple counters.

 

See backtrace() below.

 

DBGvpp# 

Thread 1 "vpp_main" received signal SIGSEGV, Segmentation fault.

0x76c4b451 in vlib_increment_combined_counter (cm=0x7fffb4f3b740,
thread_index=0, 

index=3, n_packets=256, n_bytes=36608) at
/home/hemant/int2/vpp/src/vlib/counter.h:248

248  /home/hemant/int2/vpp/src/vlib/counter.h: No such file or directory.

(gdb) bt

#0  0x76c4b451 in vlib_increment_combined_counter
(cm=0x7fffb4f3b740, 

thread_index=0, index=3, n_packets=256, n_bytes=36608)

at /home/hemant/int2/vpp/src/vlib/counter.h:248

#1  0x76c4adab in vnet_interface_output_node_inline (

vm=0x762a0d40 , node=0x7fffb7c7af40,
frame=0x7fffb7d66f80, 

vnm=0x77b4b400 , hi=0x7fffb4b6a980, do_tx_offloads=0)

at /home/hemant/int2/vpp/src/vnet/interface_output.c:418

#2  0x76c47064 in vnet_interface_output_node (

vm=0x762a0d40 , node=0x7fffb7c7af40,
frame=0x7fffb7d66f80)

at /home/hemant/int2/vpp/src/vnet/interface_output.c:498

#3  0x75fe2378 in dispatch_node (vm=0x762a0d40
, 

node=0x7fffb7c7af40, type=VLIB_NODE_TYPE_INTERNAL, 

dispatch_state=VLIB_NODE_STATE_POLLING, frame=0x7fffb7d66f80, 

last_time_stamp=77539673133597) at
/home/hemant/int2/vpp/src/vlib/main.c:1228

#4  0x75fe2e07 in dispatch_pending_node (vm=0x762a0d40
, 

pending_frame_index=2, last_time_stamp=77539673133597)

at /home/hemant/int2/vpp/src/vlib/main.c:1389

#5  0x75fdc661 in vlib_main_or_worker_loop (vm=0x762a0d40
, 

is_main=1) at /home/hemant/int2/vpp/src/vlib/main.c:1877

#6  0x75fde3aa in vlib_main_loop (vm=0x762a0d40
)

at /home/hemant/int2/vpp/src/vlib/main.c:2005

#7  0x75fde192 in vlib_main (vm=0x762a0d40 , 

input=0x7fffa799efa8) at /home/hemant/int2/vpp/src/vlib/main.c:2251

---Type  to continue, or q  to quit---

#8  0x760695de in thread0 (arg=140737323339072)

at /home/hemant/int2/vpp/src/vlib/unix/main.c:670

#9  0x753f8058 in clib_calljmp ()

at /home/hemant/int2/vpp/src/vppinfra/longjmp.S:123

#10 0x7fffd1b0 in ?? ()

#11 0x76069124 in vlib_unix_main (argc=3, argv=0x7fffe688)

at /home/hemant/int2/vpp/src/vlib/unix/main.c:747

#12 0x00406d2f in main (argc=3, argv=0x7fffe688)

at /home/hemant/int2/vpp/src/vpp/vnet/main.c:338

(gdb)

 

Thanks,

 

Hemant

 



smime.p7s
Description: S/MIME cryptographic signature

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#19014): https://lists.fd.io/g/vpp-dev/message/19014
Mute This Topic: https://lists.fd.io/mt/81576847/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] TCP delay ACK

[liuyacan]

Hi  Florin,     Oh, I just found it by accident When I compare the packet capture result with the kernel's.Best regards,Yacan




 

 

On 3/24/2021 15:09，Florin Coras wrote： 


Hi Yacan, We don’t support delayed acks on purpose, the reasoning being that by construction we only send an ack per burst of packets received. Moreover, delayed acks tend to interfere with congestion control algorithms and rtt estimators. Out of curiosity, why are you looking into that?Regards,FlorinOn Mar 23, 2021, at 10:57 PM, liuyacan  wrote:Hi All,     It seems that  macro TCP_ALWAYS_ACK is not used now，Are there any considerations? Is there any way I can Enable delay ACK ?Best regards,yacan





-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#19013): https://lists.fd.io/g/vpp-dev/message/19013
Mute This Topic: https://lists.fd.io/mt/81570505/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

[vpp-dev] [tsc] Call for Mentors and Projects for 2021 LF Networking Mentorship Program

[Trishan de Lanerolle]

FYI:

-- Forwarded message -
From: Casey Cain 
Date: Wed, Mar 17, 2021 at 3:53 AM
Subject: [tsc] Call for Mentors and Projects for 2021 LF Networking
Mentorship Program
To:


Today, we’re excited to open the call for mentors and project proposals for
the 2021 LF Networking Mentorship Program
. This is
the first year we're bringing all community intern/mentor projects under
the LF Networking umbrella to launch with an integrated program schedule
and program operations/management. The Program is intended to provide a
formal structure to connect mentors and student developers from around the
globe to contribute their enthusiasm, time, and experience toward building
sustainable LFN communities.

 *Why should you consider mentoring for LF Networking Mentorship Program?*

   - You believe in the value of mentorship in helping new developers to
   navigate open source development, the culture, the tooling and
   infrastructure to be a productive member of the community.
   - You are passionate about teaching and guiding student developers, many
   of whom may be first time open-source contributors,
   - You are eager to bring in new perspectives, new ideas, and new talent
   into your community and projects.
   - You have a narrowly scoped mentorship project with clear learning
   objectives/outcomes and a mentee's work and potential contributions could
   add value to the project or community of which you're an active developer
   or maintainer.

 *I am interested in mentoring but how do I get started? *

   - Please submit a project proposal
    for consideration by the TSC no
   later than Monday, *April 5th*.
   - Please reference 2021 program schedule
   as you plan out your project and
   commitment.
   - Please review the project proposal guidelines
    before you
   submit your proposal

*What are the elements the LFN Mentorship Program will implement to
maximize mentor/mentee collaboration success?*

   - Mentor and mentee onboarding will be conducted at the start of the
   Program
   - Mentor and mentee guide will be available as part of the Linux
   Foundation CommunityBridge Mentorship initiative
   - Project planning that includes deliverables, milestones, and tasks
   will be completed collaboratively between the mentors and mentees during
   the first two weeks of the program and posted on the wiki to increase
   transparency and accountability.
   - Mentee presentations will be required to sharpen both presentation
   skills and to cultivate the ability to provide constructive feedback and
   critique that’s the norm in the open source community for developing
   technologies collaboratively and openly.
   - Mentee evaluation will be conducted on a regular cadence tied to
   milestone deliverable schedules to help mentors and LFN staff formalize and
   respond to progress.

While mentors will be on a voluntary basis, the hired mentees will be
eligible to receive a stipend. As an added bonus, each mentee who
successfully completes the Program will be invited and financially
sponsored by LF Networking to attend an event/conference and present their
work to the broader community (specific event TBD with exemptions allowing
for health and safety concerns.).

If you have any questions, please contact mentors...@lfnetworking.org

We look forward to your submission of a mentorship project and thank you in
advance for volunteering your time to contribute to the training of the new
talent pool in the LF Networking communities.

Best,
Casey Cain
Technical Program Manager / Community Architect
Linux Foundation
_
IRC: CaseyLF
WeChat: okaru6
Voice: +1.408.641.0193
Book a Meeting w/ Casey 



-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#19012): https://lists.fd.io/g/vpp-dev/message/19012
Mute This Topic: https://lists.fd.io/mt/81583665/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] VPP on Windows

[Benoit Ganne (bganne) via lists.fd.io]

Hi Lin,

> I am new to VPP. I am considering VPP for Windows.
> Please let me know if anyone has used VPP for Windows.

There is no chance it will work, that would require quite a lot of work plus 
increased maintenance burden.
TBH I do not think there is interest in the community for that - but maybe I'm 
wrong.

Best
ben

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#19009): https://lists.fd.io/g/vpp-dev/message/19009
Mute This Topic: https://lists.fd.io/mt/81558423/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] Requirement of predictable interface name

[chetan bhasin]

Thanks a lot Venu.

It works for me.

Regards,
Chetan



On Wed, Mar 24, 2021 at 12:21 PM Venumadhav Josyula 
wrote:

> Hi Chetan,
>
> That ability is already there right now also we are using vpp 20.09 and
> intend to go 21.x also, we do it this way
> ...
> dpdk {
> dev :19:00.1 {
> name vpp-intf19/0/1  <--- "you can give whatever you , we have
> chose bus, slot,fn appended with vpp-intf"
>  }
> ..
>
> }
>
> We doing this way and it works for us. This generation fo startup.conf
> happens to us via some our task ( unix process ). I guess that should good
> enough ?
>
> Thanks,
> Regards,
> Venu
>
>
> On Wed, 24 Mar 2021 at 11:31, chetan bhasin 
> wrote:
>
>> Hi Team,
>>
>> We have a requirement to have a predictable interface name , say
>> "device_" , instead of GigabitEthernet/TenGigabitEthernet etc
>> , so that external scripts can create vpp.conf automatically.
>>
>> One way is to modify plugins/dpdk/device/format.c code as per our
>> requirement . but we want to avoid changes in vpp code base that will make
>> vpp upgrade easy for the future.
>>  Can you please suggest any additional CLI or configuration via which we
>> could alias interface names ?
>>
>> Thanks,
>> Chetan
>>
>> 
>>
>>

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#19008): https://lists.fd.io/g/vpp-dev/message/19008
Mute This Topic: https://lists.fd.io/mt/81570537/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] TCP delay ACK

[Florin Coras]

Hi Yacan, 

We don’t support delayed acks on purpose, the reasoning being that by 
construction we only send an ack per burst of packets received. Moreover, 
delayed acks tend to interfere with congestion control algorithms and rtt 
estimators. 

Out of curiosity, why are you looking into that?

Regards,
Florin

> On Mar 23, 2021, at 10:57 PM, liuyacan  wrote:
> 
> Hi All,
> 
>  It seems that  macro TCP_ALWAYS_ACK is not used now，Are there any 
> considerations? Is there any way I can Enable delay ACK ?
> 
> Best regards,
> yacan
> 
> 


-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#19007): https://lists.fd.io/g/vpp-dev/message/19007
Mute This Topic: https://lists.fd.io/mt/81570505/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] Requirement of predictable interface name

[Venumadhav Josyula]

Hi Chetan,

That ability is already there right now also we are using vpp 20.09 and
intend to go 21.x also, we do it this way
...
dpdk {
dev :19:00.1 {
name vpp-intf19/0/1  <--- "you can give whatever you , we have
chose bus, slot,fn appended with vpp-intf"
 }
..

}

We doing this way and it works for us. This generation fo startup.conf
happens to us via some our task ( unix process ). I guess that should good
enough ?

Thanks,
Regards,
Venu


On Wed, 24 Mar 2021 at 11:31, chetan bhasin 
wrote:

> Hi Team,
>
> We have a requirement to have a predictable interface name , say
> "device_" , instead of GigabitEthernet/TenGigabitEthernet etc
> , so that external scripts can create vpp.conf automatically.
>
> One way is to modify plugins/dpdk/device/format.c code as per our
> requirement . but we want to avoid changes in vpp code base that will make
> vpp upgrade easy for the future.
>  Can you please suggest any additional CLI or configuration via which we
> could alias interface names ?
>
> Thanks,
> Chetan
>
> 
>
>

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#19006): https://lists.fd.io/g/vpp-dev/message/19006
Mute This Topic: https://lists.fd.io/mt/81570537/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

[vpp-dev] Requirement of predictable interface name

[chetan bhasin]

Hi Team,

We have a requirement to have a predictable interface name , say
"device_" , instead of GigabitEthernet/TenGigabitEthernet etc
, so that external scripts can create vpp.conf automatically.

One way is to modify plugins/dpdk/device/format.c code as per our
requirement . but we want to avoid changes in vpp code base that will make
vpp upgrade easy for the future.
 Can you please suggest any additional CLI or configuration via which we
could alias interface names ?

Thanks,
Chetan

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#19005): https://lists.fd.io/g/vpp-dev/message/19005
Mute This Topic: https://lists.fd.io/mt/81570537/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-